© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Gemma-2-27B
    5. 34-GEMMASCOPE-RES-131K
    7. 66279
    Prev
    Next
    INDEX
    Explanations

    code project file headers

    np_acts-logits-general · gemini-2.5-flash-lite

    The neuron is highly sensitive to the very first token of a new line or code block, i.e. it marks line-initial words.

    oai_token-act-pair · o4-miniTriggered by @jyhe0408
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    google/gemma-scope-27b-pt-res/layer_34/width_131k
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     इसे
    -0.80
    ╹
    -0.79
    thirds
    -0.75
    VolleyError
    -0.75
     олив
    -0.75
    ohistochemical
    -0.75
     baf
    -0.74
    jsx
    -0.74
     LECT
    -0.73
    ными
    -0.73
    POSITIVE LOGITS
    ЕНЬ
    0.88
    借助
    0.87
     bonnes
    0.82
     (§
    0.81
     accuses
    0.80
     fragen
    0.79
     démocr
    0.76
    wien
    0.76
     VERY
    0.74
    limsy
    0.74
    Activations Density 0.006%

    No Known Activations